Reference Conditions for Old-Growth
Redwood Restoration on Alluvial Flats
Christa M. Dagley1 and John-Pascal Berrill1
Abstract
We quantified structural attributes in three alluvial flat old-growth coast redwood stands. Tree
size parameters and occurrences of distinctive features (e.g., burls, goose pens) were similar
between stands. Occurrence of distinctive features was greater among larger trees. Tree sizefrequency distributions conformed to a reverse-J diameter distribution. The range of tree sizes
was similar between study sites. Redwood density ranged from 118 to 148 trees ha-1 and
upper canopy tree density ranged from 45 to 74 trees ha-1. Crown ratio was similar across
study sites with an overall mean of 64.3 percent, except that crown ratio of the largest trees
was lower at the site with the highest growing space occupancy. The percentage of plot area
in canopy gaps ranged from 17 to 25 percent. Seedling regeneration was no more frequent
beneath canopy gaps. These and other results describe structure in old-growth redwood forests
and can serve as reference conditions for old forest restoration on alluvial flats.
Key words: canopy gaps, regeneration, Sequoia sempervirens, stand structure
Introduction
Old-growth redwood (Sequoia sempervirens) forests are structurally diverse,
having been shaped over centuries by a wide range of forces (Lorimer et al.2009).
Trees with diameters of 2 to 4 m and heights of 60 to 100 m are common. The desire
to restore some of these features in managed forests has prompted application of
various silvicultural practices, yet to date, little attention has been directed at
developing a detailed description of structural parameters and distinctive features
found in old-growth redwood forests that describe the historic condition and could
serve as reference conditions for restoration.
Quantification of old-growth forest structure supports restoration efforts by
providing reference conditions that could serve as targets for management activities
(Harrod et al. 1999, SERISPWG 2004). The objective of this study was to identify
structural parameters and distinctive features that were characteristic of alluvial flat
old-growth redwood stands and may thus serve as targets for restoration of old forest
structures. We hypothesized that: (1) multivariate analysis would not successfully
discriminate between study sites on the basis of tree size and crown parameters; (2)
seedling regeneration was no more frequent directly beneath canopy gaps than
beneath tree crowns; and (3) the occurrence of distinctive features (e.g., epicormic
sprouting) on live trees differed between sites, and was greater among larger trees. To
test these hypotheses, calculate canopy gap areas and canopy volume, and summarize
total live stem biomass, coarse woody debris, and regeneration data, we made
detailed measurements of various tree size and crown size parameters, mapped tree
locations, and recorded instances of distinctive features on individual redwood trees.
1
Research Associate and Assistant Professor, respectively, Department of Forestry and Wildland
Resources, Humboldt State University, Arcata, CA 95521 (cd104@humboldt.edu).
251
GENERAL TECHNICAL REPORT PSW-GTR-238
Methods
Three study sites were selected—from the few larger remaining old growth
groves on alluvial flats in northern California—to each represent different conditions
and locations. Two of the study sites, named Children’s and Rockefeller, were
located at Humboldt Redwoods State Park in the southern part of Humboldt County
(N 40°18.5’, W 123°54.4’). The study site at Children’s Forest in HRSP was located
along the south fork of the Eel River adjacent to Myers Flat. Children’s differed from
the other two study sites in that a wildfire occurred there 1.5 years prior to study
installation. The Rockefeller site in HRSP was approximately 14 km NW of the
Children’s site and located along the small perennial Bull Creek. Mean annual
temperature for the HRSP area is 12.6 °C and mean annual precipitation is 123 cm.
The third study site, named Armstrong, was located alongside a seasonally dry creek
at Armstrong Redwoods State Preserve (ARSP) in the Russian River region of
Sonoma County (N 38°53’, W 123°0.6’). Mean annual temperature for ARSP is 13.9
°C and mean annual precipitation totals 104 cm.
Plants common to these redwood-dominated sites included redwood sorrel
(Oxalis oregano) and western sword fern (Polystichum munitum). Hardwoods such as
tanoak (Notholithocarpus densiflorus), bigleaf maple (Acer macrophyllum), and
California bay (Umbellularia californica) were found in small numbers in openings
or near streams.
At each site, a 1 ha sample plot (100 m x 100 m) was established with the
objective of quantifying the horizontal and vertical structure. All trees ≥ 15 cm
diameter at 1.37 m breast height (dbh) were mapped using a survey laser mounted on
a tripod. Species, dbh, and location (e.g., azimuth and horizontal distance to the
middle of each tree from reference points) were recorded. To assure mapping was
accurate, the angle and distance from plot center to each reference point was recorded
and cross-checked with tree locations taken from both plot center and the respective
reference point. Additionally, all trees ≥ 15 cm dbh were measured for total height,
live crown base height, and crown radius. Live crown base height was defined as the
height of the lowest major branch that formed part of the main canopy. The number
of crown radius measurements varied with complexity of crown shape. A minimum
of four crown radius measurements were taken, each in a cardinal direction, for small
trees with circular crowns. A maximum of eight measurements were taken for large
trees with irregular crowns. Crown class, crown shape, and crown fullness estimates
were recorded for each tree. The three-dimensional shape of individual tree crowns
was most often described as a parabola; however, some were described as a cone,
cylinder, or diamond. Crown fullness was an ocular estimation of the percentage of
the live crown occupied by branches and foliage (e.g., 50, 66, 75 percent). Because
older trees often have discontinuous crowns, incorporating the “fullness” variable
into the crown volume equation provided a more accurate estimate of actual space
occupied by live tree crowns.
For each redwood tree, we noted presence or absence of the following distinctive
features: burls, epicormic sprouting, goose pens, and reiterations. A goose pen was
defined as a large fire-created basal hollow greater than 30 or 60 cm tall for trees less
than or greater than 1 m dbh, respectively. A reiteration was defined as an erect
sprout or branch found on an existing tree but supporting its own network of
horizontally oriented branches. Three 100 m north-south transects were established in
252
Reference Conditions for Old-Growth Redwood Restoration on Alluvial Flats
each 1 ha plot with the objective of quantifying regeneration, understory vegetation,
and groundcover. Height, location, species, and regeneration source (i.e., sprout
versus seedling) were recorded for all trees < 15 cm dbh in a continuous 4 m wide
belt centered on each transect line. Percent ground cover values were obtained by
dividing the entire length of each transect line into one centimeter sections and
recording species of shrubs, herbaceous vegetation, bareground, and litter covering
ground directly below the transect line. Down logs ≥ 30 cm diameter and ≥ 2 m
length were mapped and measured to determine the volume, mass, and percent cover
of large down wood at Children’s and Rockefeller. Diameter at the midpoint of each
length of down log, log length, and location were recorded. Only the portions of logs
falling inside the plot were measured. Down logs at Armstrong were not measured
due to a history of removal of down wood from the site. Standing dead trees (snags)
≥ 15 cm dbh were mapped and measured for dbh at the three study sites.
Data for live trees with dbh ≥ 15 cm were summarized to give per-hectare stem
density, basal area, and stem volume, stand density index (Reineke 1933), and
maximum tree height for each species at the three study sites. Stem volume (V) was
calculated using a simple conic shape: V=π(0.5 dbh)2h/3 where dbh = dbh in m and h
= total tree height in m. Stand density index (SDI) was calculated as a summation of
individual tree values because the dbh data were not normally distributed:
SDI=∑(0.04dbhi)a where dbhi = dbh in cm of the ith tree in the plot, and a = 1.605
(Long and Daniel 1990, Shaw 2000). Trees were grouped into 30 cm dbh size classes
to give size-frequency distributions for each site. Individual tree size parameters of
dbh, height, crown radius, crown volume, and crown ratio (ratio of crown length to
total tree height) were analyzed using canonical discriminant analysis. The objective
was to identify individual parameters or groups of parameters that differed least
between sites, assuming these ‘commonalities’ would be more useful targets for
restoration than parameters that differed between sites. Prior to analysis, a
logarithmic transformation of dbh was applied to achieve normality of the residual
variances.
After arc-sine transformation, average crown ratio was calculated for redwood
trees in each 30 cm dbh size class. Average crown ratios for each size class were
subjected to an ANOVA to test for significant differences (α = 0.05) between study
sites. The length of clear bole was defined as the portion of the tree stem not
occupied by the canopy (equivalent to live crown base height). Clear bole length was
summarized by 30 cm dbh size classes to further describe the vertical structure of
standing trees at each site. Crown volume for each tree was calculated using crown
length, radius, shape, and fullness. Summing crown volume data for all trees in the 1ha plots allowed for a site-by-site comparison of total canopy volume. To determine
the vertical distribution of crowns and crown volume at each site, individual tree
crown volume was separated into layers at 10-m height intervals and summed for
each 10 m layer.
ArcGIS ArcMap (ESRI) was used to delineate and calculate plot area not covered
by the downward projection of tree crowns. Crown canopy maps of each site were
created using stem locations and crown radius measurements. Spatial coordinates of
crown extent were created for each crown radius measurement and then connected
for each tree to create polygons representing tree crown extent. Polygon boundaries
were smoothed using a t-spline. To obtain an accurate estimate of gap area within the
1 ha plots, portions of the crown of trees surrounding the plot that extended inside
253
GENERAL TECHNICAL REPORT PSW-GTR-238
plot boundaries were included in the crown canopy maps. This necessitated mapping
and measurement of crown radii for trees with crowns encroaching on the plot area.
Crown canopy maps were used to examine the relationship between canopy gaps and
regeneration.
Location coordinate data for regeneration in each 4-m wide belt transect were
imported into the crown canopy maps for each site. Each record of regeneration (all
trees <15 cm dbh) was coded as occurring directly beneath a canopy gap or beneath
an opening in the canopy. Binomial proportions tests were used to test for differences
in occurrence of regeneration beneath tree crowns versus regeneration under canopy
gaps, by species type (i.e., redwood or hardwood) and source of regeneration (i.e.,
sprout or seedling). Count data for regeneration under canopy gaps or tree crowns
were each divided by the proportions of transect area either under canopy gaps or
beneath tree crowns, respectively, making count data comparable when different
proportions of transect area were located directly beneath the canopy and under
canopy gaps.
Logistic models were developed to predict probability of occurrence of burls,
epicormic sprouting, goose pens, and reiterations on individual redwood trees as a
function of tree-size parameters. Dummy variables for study site were also included
to test for significant differences in presence-absence of these distinctive features
between the three study sites. Model goodness-of-fit in terms of -2 Log Likelihood
was compared between models to identify the tree size parameter (i.e., dbh, height,
live crown base height, crown length, crown radius) most strongly associated with
probability of occurrence of distinctive features on redwood trees. Data were
analyzed using SAS statistical analysis software (SAS Institute 2004).
Results
Redwood density ranged from 118 to 148 trees ha-1 and density of all live stems
ranged from 118 to 183 trees ha-1 for the three study sites (table 1). The two sites at
Humboldt Redwoods State Park (Children’s and Rockefeller) were almost
completely comprised of redwood. In contrast, the site at Armstrong Redwood State
Park (Armstrong) contained a mix of hardwoods. Hardwood crowns had not attained
upper canopy status. The distribution of stems by 30 cm dbh class indicated that
Armstrong’s higher stem density was mostly in the small size classes and because of
the hardwood presence. The frequency distribution of redwood tree diameters was a
reverse-J shape at the three study sites. The two northern sites (Children’s and
Rockefeller) contained larger diameter trees while a narrower distribution of
diameters at Armstrong was observed (fig. 1). Mean dbh was similar at Armstrong
and Children’s (1.10 and 1.08 m, respectively). Mean dbh at Rockefeller was 1.45 m.
254
Reference Conditions for Old-Growth Redwood Restoration on Alluvial Flats
Table 1—Species composition and stem density per hectare by crown class for all live trees ≥
15 cm dbh in the 1-ha plot at the three alluvial flat old-growth redwood study sites.
Site
Armstrong
Species
Dominant
S. sempervirens
47
A. macrophyllum
-N. densiflorus
-U. californica
-Total
56
Children’s S. sempervirens
13
U. californica
-Total
14
Rockefeller S. sempervirens
29
Codom.
27
---27
33
-33
16
Intermediate Suppressed
27
47
4
2
5
2
19
3
46
54
45
37
2
1
46
38
44
29
Total
148
6
7
22
183
128
3
131
118
Figure 1—Diameter distribution of live trees ≥ 15 cm dbh for 1-ha plots at the
three alluvial flat old-growth redwood study sites.
Structural attributes of redwood basal area, canopy volume, and total standing
stem volume tended to be similar at Children’s and Armstrong study sites, and
highest at Rockefeller (table 2). Among these structural attributes, total standing stem
volume differed most between study sites. Basal area and total standing stem volume
at Rockefeller were at least 30 and 37 percent greater, respectively, than at Children’s
and Armstrong. Similarly, redwood SDI was greater at Rockefeller than at Children’s
and Armstrong (metric SDI: 2529, 1884, 2057, respectively). Maximum height was
similar at the two northern sites and taller compared to Armstrong. Total canopy
volume for the three sites was similar in overall value but was distributed differently
along the vertical profile (table 2, fig. 2). Peak canopy volume values differed
between the three study sites; Armstrong peaked at a lower height (40 m) than the
two northern sites, Children’s peaked at 50 m, and Rockefeller had the most crown
volume at 60 m above ground.
255
GENERAL TECHNICAL REPORT PSW-GTR-238
Table 2—Basal area (BA), canopy volume, total standing stem volume, and height of the
tallest tree (Max. Ht.) for live trees ≥ 15 cm dbh in the 1-ha plot at the three alluvial flat oldgrowth redwood study sites.
Site
Species
Armstrong
S. sempervirens
A. macrophyllum
N. densiflorus
U. californica
Total
S. sempervirens
U. californica
Total
S. sempervirens
Children’s
Rockefeller
BA
(m2 ha-1)
228.3
0.3
0.3
4.8
233.7
225.1
0.8
225.9
307.8
Canopy volume
(m3 ha-1)
168,281
838
588
15,363
185,070
171,806
1,186
172,992
174,588
Stem volume Max. Ht.
(m3 ha-1)
(m)
5,542
95
1.7
24
1.3
18
60.2
49
5,605
-6,732
105
15.9
59
6,748
-9,272
107
Canonical discriminant analysis revealed significant structural differences
between the three sites. Armstrong and Children’s sites were found to be the most
different (P < 0.0001) while the two northern sites, Children’s and Rockefeller, were
found to be the least different (P = 0.007). The analysis discriminated between the
sites predominantly on the basis of dbh and height, suggesting that height to diameter
ratio was the most important structural difference between sites. The analysis also
revealed different tree diameter – crown volume relations between sites. However,
modest allocation rates (42 to 57 percent) indicated that the sites could not be
separated easily by tree size parameters.
Figure 2— Canopy volume distribution by 10-m height interval along vertical
profile for each alluvial flat old-growth redwood study site, based on tree crown
volume estimates for redwood stems ≥ 15 cm dbh.
Long crowns were a common feature at the three study sites (table 3). The
overall mean crown ratio for the three study sites was 64.6 percent. An F-test failed
to detect significant differences in mean crown ratio for each 30-cm dbh size class
between the three study sites (P = 0.14). However, differences in crown ratio
256
Reference Conditions for Old-Growth Redwood Restoration on Alluvial Flats
between sites were detected in an F-test of tree crown ratio data for large trees (i.e.,
trees > 1.5 m dbh; P = 0.001). Further analysis of data for large trees detected no
significant difference in crown ratio between Children’s and Armstrong sites (P =
0.92), whereas crown ratios at these two sites differed significantly from the
Rockefeller site (P ≤ 0.005) where the larger trees had longer clear boles and shorter
crowns (table 3).
Table 3—Average crown ratio (CR) and length of clear bole (CB) by 30 cm dbh size class for
all redwood trees ≥ 15 cm dbh at the three alluvial flat old-growth redwood study sites.
Dbh class
midpoint (cm)
30
60
90
120
150
180
210
240
270
300
330
360
390
420
Mean
n
45
24
12
14
11
7
15
11
6
3
0
0
0
0
--
Armstrong
CR
CB (m)
.61
6
.71
9
.69
16
.65
22
.64
24
.74
21
.67
23
.55
35
.61
32
.65
30
--------.65
--
n
51
18
8
15
8
8
2
3
3
2
1
5
3
1
--
Children’s
CR
CB (m)
.62
8
.70
13
.69
21
.66
24
.66
27
.61
34
.74
22
.68
30
.64
34
.65
35
.47
47
.61
39
.73
27
.84
15
.65
--
n
31
11
11
7
12
11
6
5
5
6
6
3
2
2
--
Rockefeller
CR
CB (m)
.68
6
.76
9
.69
17
.67
23
.69
22
.58
34
.64
32
.50
44
.59
40
.58
41
.50
47
.50
47
.62
38
.54
46
.64
--
The sum of canopy gap areas as a percentage of total plot area was similar at the
three sites, ranging from 17 to 25 percent. Gap shape and size were variable (fig. 3).
For gaps contained within plot boundaries there was an approximate reverse-J
distribution in gap sizes. Rockefeller had one large gap (>1272 m2) which extended
beyond the plot boundary and accounted for 50 percent of the total gap area within
the plot.
Figure 3—Canopy map of tree crown projections in 1-ha plots at (a) Armstrong; (b)
Children’s; and (c) Rockefeller alluvial flat old-growth redwood study sites.
257
GENERAL TECHNICAL REPORT PSW-GTR-238
Regeneration density and species composition differed between study sites (table
4). Hardwood trees < 15 cm dbh were found at Armstrong and Rockefeller.
Armstrong contained a mix of hardwoods including tanoak, California bay, and red
alder (Alnus rubra). In contrast, Rockefeller contained only one hardwood species,
tanoak. Redwood sprouts and seedlings were present at all three sites. Binomial
proportions tests did not detect differences between the occurrence of redwood
sprouts or seedlings occurring in canopy gaps and redwood regeneration beneath the
canopy (P ≥ 0.31), nor between hardwood sprouts or individuals occurring in gaps or
beneath the canopy (P ≥ 0.12). Most redwood regeneration found along transects was
in the form of basal sprouts associated with an existing root system.
Table 4—Regeneration density per hectare for all trees < 15 cm dbh in three 400 m2 belt
transects at each alluvial flat old-growth redwood study site.
Site
Armstrong
Children’s
Rockefeller
Species
Sequoia sempervirens
Hardwoods
Sequoia sempervirens
Sequoia sempervirens
Hardwoods
Sprouts
1125
404
1538
158
483
Seedlings
48
538
25
150
517
Total
1173
942
1563
308
1000
The main constituents of the forest floor were similar among sites, with redwood
sorrel, sword fern, and litter most prevalent. However, the proportions of these
elements differed between sites. Rockefeller contained almost an even mix of litter,
redwood sorrel, and sword fern (35, 21, and 32 percent, respectively). Children’s
groundcover was almost a complete blanket of redwood sorrel (76 percent), probably
as a result of the fire. Armstrong’s ground cover was dominated by litter (54 percent)
followed by redwood sorrel (30 percent), and sword fern (6 percent). Consistent with
the higher tree species diversity found at Armstrong, the understory was also most
diverse and included fragrant bedstraw (Galium triflorum), trail marker
(Adenocaulon bicolor), fairybell (Disporum hookeri), and wood rose (Rosa
gymnocarpa).
Epicormic sprouting was the most common distinctive feature at each site. Burls,
epicormic sprouting, goose pens, and reiterations were more common among larger
redwood trees (fig. 4). Most redwood trees had one type of distinctive feature and
very few trees had all of these features. The natural logarithm of tree dbh was a
stronger predictor of the probability of occurrence of distinctive features than either
tree height, live crown base height, crown length, or crown radius. The probability of
occurrence of epicormic sprouting was greater at Children’s and lower at Armstrong
when compared with incidence of epicormic sprouting for any given tree size at
Rockefeller. Logistic model coefficients indicated that redwood trees at Children’s
were less likely to have a goose pen than trees of equivalent dbh at either Armstrong
or Rockefeller (table 5).
258
Reference Conditions for Old-Growth Redwood Restoration on Alluvial Flats
Figure 4—Percentage frequency distribution of redwood trees having a distinctive
feature by 30 cm dbh class, based on pooled data from the three alluvial flat oldgrowth redwood study sites.
Table 5—Coefficients and fit statistics for logistic model of the presence of burls, epicormic
sprouts, goose pens, and reiterations as a function of the natural logarithm of dbh for alluvial
flat old-growth redwood trees ≥15 cm dbh (n=394). Dummy variables for study sites were
included when differences detected between sites (α=0.05).
Distinctive Feature
Burls
(Model -2LL = 346.5)
Epicormic sprouts
(Model -2LL = 438.9)
Goose pens
(Model -2LL = 225.8)
Reiterations
(Model -2LL = 327.1)
Parameter
Intercept
Ln dbh
Intercept
Ln dbh
Armstrong
Children’s
Rockefeller
Intercept
Ln dbh
Armstrong
Children’s
Rockefeller
Intercept
Ln dbh
Coefficient
1.0617
-1.7086
-1.0899
-0.3871
-0.7871
0.7660
0.0000
2.4330
-1.7091
0.2623
-0.8022
0.0000
1.6237
-0.8489
-2LL = -2 Log Likelihood, measure of goodness of fit.
s.e. = standard error for coefficient.
s.e.a
0.15
0.20
0.13
0.13
0.16
0.19
-0.23
0.28
0.25
0.24
-0.15
0.17
Wald’s χ2
52.70
76.08
70.94
9.43
25.48
16.47
-107.86
36.91
1.12
11.31
-123.64
24.51
Pr > χ2
<0.0001
<0.0001
<0.0001
0.0021
<0.0001
<0.0001
-<0.0001
<0.0001
0.2894
0.0008
-<0.0001
<0.0001
a
The number of large fallen logs was similar at Children’s and Rockefeller study
sites. Logs at Rockefeller were slightly longer and larger in diameter which resulted
in almost twice the amount of volume and area covered by logs at Rockefeller (table
6). There appeared to be no trend in the direction of tree fall and most of the logs
were found to be scattered widely across each site. One exception occurred at
Rockefeller. The eastern side of the plot contained a large gap which was created by
three large fallen trees that fell in the easterly direction.
259
GENERAL TECHNICAL REPORT PSW-GTR-238
Table 6—Coarse woody debris characteristics for Children’s and Rockefeller alluvial flat
old-growth redwood study sites. Minimum dimensions of logs sampled were 2 m length and
30 cm midpoint diameter. Standard error for mean log diameter shown in parentheses.
Site
Children’s
Rockefeller
a
b
Density
(logs ha-1)
14
19
Mean diameter
(m)
1.09 (0.12)
1.25 (0.17)
Volume
(m3 ha-1)
570
1072
Cover a
(%)
3.7
6.8
Mass b
(tons ha-1)
114
214
Cover represents percent ground area covered by downward projection of logs.
Mass estimated as volume by an averaged wood density from Bingham and Sawyer (1988).
Standing dead trees (snags) ≥ 15 cm dbh appeared to be exclusively redwood at
the three sites. The two snags within the plot at Armstrong were 1.5 and 1.6 m dbh.
The two snags at Rockefeller were 0.7 and 3.5 m dbh. The eight snags at Children’s
included four small dead-standing redwood trees (dbh < 40 cm) that appeared to have
sustained severe fire damage and died recently. The other four snags at Children’s
ranged from 0.9 to 2.2 m dbh.
Discussion
Our work provides an important description of structural characteristics in
alluvial flat old-growth redwood forests. Density, diameter distribution, crown ratio,
and total canopy gap area were most similar among study sites. As such, they may
serve as key reference conditions for future restoration treatments (table 7).
Table 7—Reference conditions for alluvial flat old-growth redwood forests.
Condition
Stand densitya
Upper canopy tree densityb
Crown ratio
Canopy gap area
Coarse woody debrisc
Snags
Spatial pattern
a
Description
Redwood 118-148 trees ha-1
45-74 trees ha-1; mean dbh 2.1 m; range 0.55-4.27 m dbh
Trees > 1.5 m dbh: 0.56-0.64; trees ≤ 1.5 m dbh: 0.65
17-25% total area in gaps
Density 14-19 logs ha-1; mean midpoint diameter 1.2 m; mean
length 26.6 m; cover 3.7-6.8%; mass 114-214 tons ha-1
2-4 dead standing trees ha-1; range 0.9-3.5 m dbh
Random pattern for redwood > 1.5m dbh (Dagley 2008)
Trees ≥ 15 cm dbh.
Dominant and codominant trees.
c
Logs ≥ 30 cm midpoint diameter and 2 m length.
b
Similarities between our data and published density data suggest both overall
redwood density and upper canopy density can be useful reference points in old
forest restoration efforts. Van Pelt and Franklin (2000) reported a main canopy
density of 46 trees ha-1. Sugihara (1992) reported an overall redwood density of 182
trees ha-1 for trees > 10 cm dbh and a density of 67 trees ha-1 for redwood trees > 1.5
m dbh. A redwood density of 107 trees ha-1 for trees > 10 cm dbh was reported for a
stand located in Prairie Creek Redwoods State Park (Sawyer and others 2000).
Fujimori (1977) reported 66 trees ha-1 for trees > 1 m dbh.
Stand density index (SDI) is a measure of growing space occupancy, and is
useful for quantifying relative density across a wide variety of stand conditions (Long
and Daniel 1990). Reineke (1933) reported that the maximum SDI for redwood is
260
Reference Conditions for Old-Growth Redwood Restoration on Alluvial Flats
approximately 2500 (equivalent to an SDI of 1000 in English units) based on data
from even-aged second growth stands. The SDI was highest at Rockefeller, and
slightly exceeded the maximum reported by Reineke (1933). This could explain the
lower crown ratios among larger trees (table 3) and lower regeneration density (table
4) at Rockefeller.
Old-growth redwood tree crowns occupy an immense amount of space. Van Pelt
and Franklin (2000) reported a canopy volume of 230,100 m3 ha-1 with the maximum
occurring at a height of 50 m for a stand on the alluvial flats in Humboldt Redwoods
State Park. This value exceeded our estimates (172,992 to 185,070 m3 ha-1; table 2)
but was based on a conic representation of tree crown volume. We generated more
realistic estimates by describing the shape and fullness of individual tree crowns.
Crown shape impacts volume estimation and placement of crown volume in the
vertical profile (fig. 2). For example, the volume of a parabola is 50 percent greater
than that of a cone. When compared with crown volume estimates obtained using a
cone shape, canopy volume estimates from this study were 10 m higher in the vertical
profile when shape was taken into account. The shape was most often described as a
parabola and the average crown fullness was 60 percent.
Total canopy gap area ranged from 17 to 25 percent between sites. These values
closely resembled reported values from other old-growth stands ranging from 18 to
20 percent (Busing and Fujimori 2002, Sugihara 1992, Van Pelt and Franklin 2000).
Gap size and shape were variable, suggesting that treefall events involved individual
trees or groups of trees.
A preference to establish under canopy gaps was not detected among redwood
regeneration. Hunter (1995) found the influence of canopy gaps to only account for
4.6 percent of the variation in understory light levels in a mixed evergreen forest
(containing redwood) in Northern California. He proposed that several factors limit
the effects of canopy gaps on the understory: temperate latitude placing the sun at
lower angles, a dry summer season restricting growth when light is at its highest
angle in the sky, a tall canopy, and small gap diameters. Van Pelt and Franklin (2000)
found no relationship between understory tree location and canopy gaps.
Exposure to wind and windstorm events likely affects stem form. Results from
the multivariate discrimant analysis revealed differences in tree height-diameter
relations between study sites. Our use of a simple conic shape to estimate stem
volume is repeatable, but highlights need for reliable predictive volume and taper
equations. Existing equations predict merchantable volume above a tall stump to
large top diameters, not total stem volume. Differences in stem taper and incidences
of top breakage and reiteration will complicate the task of modeling stem volume and
whole-tree biomass: a priority for future research.
Our characterization of coarse woody debris highlighted differences between
sites (table 6). These differences may in part be due to stand history. Over time,
down logs may have been removed for use or aesthetics. Yet, the lowest published
estimates remain two to five times greater than those reported for most other
temperate or tropical forests (Franklin and Waring 1980). Our estimates of 114 and
214 metric tons ha-1 for Children’s and Rockefeller, respectively, fall within the range
of values reported for alluvial flats (Busing and Fujimori 2005, Sugihara 1992).
The study sites at Armstrong, Children’s, and Rockefeller were selected to
261
GENERAL TECHNICAL REPORT PSW-GTR-238
represent alluvial flat old-growth redwood forests with different site characteristics
and/or recent disturbance histories. Our data and analyses revealed that each stand
had unique attributes, but also shared features with the other study sites, suggesting
that these commonalities could represent general reference conditions or “targets” for
future restoration efforts (table 7). For example, managers seeking to accelerate
development of old-growth characteristics in a young stand might begin by
identifying approximately 50 to 80 overstory trees ha-1 in a random spatial
arrangement, and releasing these trees from competition. Retaining more trees would
allow for artificial creation of 2 to 4 snags ha-1 in the absence of natural mortality.
Concurrently identifying one or more open areas to serve as large canopy gaps
between overstory tree crowns might preclude future conflict between overstory tree
density and canopy gap area requirements. Less regard, at least initially, might be
paid to structural parameters or distinctive characteristics that differed between the
three reference sites or that correlated with tree size. When combined with
descriptions of the spatial pattern of tree locations within stands (Dagley 2008), this
work represents a detailed quantitative description of the three dimensional structure
and complexity found in old-growth redwood forests on alluvial flats.
Acknowledgments
The project was funded by Save-the-Redwoods League and The Center for
Forest Disturbance Science, Southern Research Station, USDA Forest Service.
References
Bingham, B.B.; Sawyer, J.O. 1988. Volume and mass of decaying logs in an upland oldgrowth redwood forest. Canadian Journal of Forest Research 18: 1649-1651.
Busing, R.T.; Fujimori, T. 2002. Dynamics of composition and structure in an old Sequoia
sempervirens forest. Journal of Vegetation Science 13: 785-792.
Busing, R.T.; Fujimori, T. 2005. Biomass, production, and woody detritus in an old coast
redwood (Sequoia sempervirens) forest. Plant Ecology 177: 177-188.
Dagley, C.M. 2008. Spatial pattern of coast redwood in three alluvial flat old-growth
forests in northern California. Forest Science 54: 294-302.
Franklin, J.F.; Waring, R.H. 1980. Distinctive features of the northwestern coniferous
forest: development, structure, and function. In: Waring, R.H., editor. Forests: fresh
perspectives from ecosystem analysis. Annual Biology Colloquium, 1979. Corvallis,
OR: Oregon State University: 59-86.
Fujimori, T. 1977. Stem biomass and structure of a mature Sequoia sempervirens stand
on the Pacific Coast of northern California. Journal of the Japanese Forest Society
59: 431-441.
Harrod, R.J.; McRae, B.H.; Hartl, W.E. 1999. Historical stand reconstruction in ponderosa
pine forests to guide silvicultural prescriptions. Forest Ecology and Management
114: 433-446.
Hunter, J.C. 1995. Architecture, understory light environments and stand dynamics in
northern California’s mixed evergreen forests. Davis, CA: University of CaliforniaDavis. Ph.D. dissertation. 105 p.
262
Reference Conditions for Old-Growth Redwood Restoration on Alluvial Flats
Long, J.N.; Daniel, T.W. 1990. Assessment of growing stock in uneven-aged stands.
Western Journal of Applied Forestry 5: 93-96.
Lorimer, C.G.; Porter, D.J.; Madej, M.A.; Stuart, J.D.; Veirs, S.D. Jr.; Norman, S.P.; O’Hara,
K.L.; Libby, W.J. 2009. Presettlement and modern disturbance regimes in coast
redwood forests: implications for the conservation of old-growth stands. Forest
Ecology and Management 258: 1038-1054.
Reineke, L.H. 1933. Perfecting a stand-density index for even-aged forests. Journal of
Agricultural Research 46: 627-638.
SAS Institute Inc. 2004. SAS/STAT 9.1 User’s Guide, Vols. 1-7. First Edition. Cary, NC:
SAS Institute Inc.
Sawyer, J.O.; Sillett, S.C.; Popenoe, J.H.; LaBlanca, A.; Sholars, T.; Largent, D.L.; Euphrat,
F.; Noss, R.F.; Van Pelt, R. 2000. Characteristics of redwood forests. In: Noss, R.F.,
editor. The redwood forest: history, ecology, and conservation of the coast redwoods.
Washington, DC: Island Press: 41-79.
Shaw, J.D. 2000. Application of stand density index to irregularly structured stands.
Western Journal of Applied Forestry 15: 40-42.
Society for Ecological Restoration International Science & Policy Working Group. 2004. The
SER international primer on ecological restoration. www.ser.org & Tucson: Society
for Ecological Restoration International. 15 p.
Sugihara, N.G. 1992. The role of treefall gaps and fallen trees in the dynamics of old
growth coast redwood (Sequoia sempervirens [D. Don] Endl.) forests. Berkeley, CA;
University of California-Berkeley. Ph.D. dissertation. 189 p.
Van Pelt, R.; Franklin, J.F. 2000. Influence of canopy structure on the understory
environment in tall, old-growth conifer forests. Canadian Journal of Forest Research
30: 1231-1245.
263